integrated and repeat testing by the method. We gratefully acknowledge the ex-

143
Clinical Chemistry 53, No. 1, 2007
integrated and repeat testing by the
same or a different glycated Hb
method.
Fig. 1. Boronate affinity chromatograms.
The x axis is time in min and the y
axis is relative absorbance at 413
nm. Nonglycated Hb elutes at
0.47 to 0.49 min and glycated Hb
elutes at 1.05 to 1.06 min (A),
index case, HbA1c 5.9%. (B), index case, HbA1c 9.4%. (C), nondiabetic sample heterozygous for
HbE. (D), nondiabetic sample heterozygous for HbCS. The HbCS
peaks were 2.4% of the total Hb
in the ␤-Thal short program. (E),
nondiabetic sample heterozygous
for HbCS. The HbCS peaks were
2.7% of the total Hb in the ␤-Thal
short program.
We gratefully acknowledge the excellent technical support provided by
the staff of the Automated Endocrinology Laboratory and the Hemoglobin Identification section of the
Special Genetics Laboratory at ARUP
Laboratories.
References
1. Tangvarasittichai O, Jeenapongsa R, Sitthiworanan C, Sanguansermsri T. Laboratory investigations of Hb Constant Spring. Clin Lab Haematol 2005;27:47–9.
2. Laig M, Pape M, Hundreiser J, Flatz G, Sanguansermsri T, Das BM, et al. The distribution of
the Hb constant spring gene in Southeast Asian
populations. Hum Genet 1990;84:188 –90.
3. Hsia YE, Ford CA, Shapiro LJ, Hunt JA, Ching NS.
Molecular screening for haemoglobin constant
spring. Lancet 1989;1:988 –91..
4. Derry S, Wood WG, Pippard M, Clegg JB, Weatherall DJ, Wickramasinghe SN, et al. Hematologic and
biosynthetic studies in homozygous hemoglobin
Constant Spring. J Clin Invest 1984;73:1673– 82.
5. Bry L, Chen PC, Sacks DB. Effects of hemoglobin
variants and chemically modified derivatives on
assays for glycohemoglobin. Clin Chem 2001;
47:153– 63.
6. Little RR, Vesper H, Rohlfing CL, Ospina, M,
Safar-Pour S, Roberts WL. Validation of the use
of boronate affinity chromatography to measure
glycated hemoglobin in the presence of HbS and
HbC traits using a mass spectrometry reference
method. Clin Chem 2005;51:264 –5.
William L. Roberts
results gave a CV of 3.4%. A patient
sample homozygous for HbA with
an HbA1c of 8.2% was analyzed 10
times in 1 assay and gave a CV of
1.1%. The fructosamine concentration of a plasma sample from the
index patient was 384 ␮mol/L (normal ⬍285 ␮mol/L), consistent with
an HbA1c result of ⬃9%.
Phenotype analysis (HPLC, Variant analyzer, Bio-Rad Laboratories,
beta-Thal short program) indicated
1.7% HbF, 4.6% P2 (HbA1c), 4.1% P3,
62% HbA0, 23% HbE, and 4.3%
HbCS. HbCS eluted in 4 distinct
peaks with retention times from 4.7–
5.2 min. When we analyzed nondiabetic samples heterozygous for HbE
(Fig. 1C) or containing HbCS (Fig. 1D
& E), the sizes of the peaks eluting
between nonglycated and glycated
Hb were consistent with the concentration of HbCS in the samples. The
presence of HbE did not produce
such a peak.
The boronate affinity column that
we used had a history of ⬃3500 injections. Analysis of the sample from
the index patients and other samples
heterozygous for HbE and HbCS on
a column from a different lot with
⬍700 injections demonstrated no
peak between nonglycated and glycated Hb. Additional testing of samples heterozygous for HbE and
HbCS on this column after 3300 and
4400 injections failed to demonstrate
a shoulder on the glycated Hb peak,
suggesting that this interference
from HbCS is specific to one lot of
columns and mobile phase.
In conclusion, the presence of
HbCS in samples for glycohemoglobin analysis by boronate affinity
chromatography has the potential to
interfere with accurate measurement
of glycated Hb. Any boronate affinity chromatogram with a shoulder on
the glycated Hb peak should lead to
careful review of how the peaks were
Department of Pathology
University of Utah
Health Sciences Center
Salt Lake City, UT and
ARUP Institute for Clinical &
Experimental Pathology
500 Chipeta Way
Salt Lake City, UT 84108
Fax 801-584-5207
e-mail [email protected].
DOI: 10.1373/clinchem.2007.078824
To Mix with Pooled Normal Plasma or
Not to Mix: A Comparative Study of 2
Approaches for Assessing Lupus
Anticoagulant Inhibitory Activity in the
Dilute Russell Viper Venom Method
To the Editor:
The recommended (1 ) step-wise approach to the study of lupus anticoag-
144
Letters
ulant (LA) begins with a prolonged
clotting time in a phospholipid-dependent clotting assay (screen), a noncorrection in a mix with pooled normal plasma (PNP), and then a
correction in the presence of increased phospholipid (confirm). Laboratories differ in their application of
these recommendations. Jacobsen et
al. (2 ) integrate screen, mix, and confirm into a single assay and analyze
the data as a lupus ratio (LR). Tripodi
et al. (3 ) propose no prior mix with
PNP even for patients on oral anticoagulant (OAC) and analyze their
data as a percentage correction. We
wished to compare these 2 approaches by using the common (4 )
dilute Russell viper venom test
(dRVVT) as the clotting assay.
Samples were collected from 500
consecutive patients (291 women,
median age 56 years) for whom LA
tests were requested during a
5-month period. Repeat tests (n ⫽ 19)
were excluded on patients as were 9
samples that contained unfractionated heparin. Of the remaining 491
patients, 85 were on OAC with international normalized ratios ranging
from 1.7 to 3.5.
Plasma from 70 healthy volunteers
(30 men and 40 women, median age
48 years) were used to establish the
cutoff values. We used PNP from a
pool of 30 similar healthy volunteers.
Blood from controls and patients
was collected and stored as described (5 ). The LR of Jacobsen et al.
(2 ) is the ratio of 2 clotting times for
1:1 mixes of patient’s plasma and
PNP, one using a dilute phospholipid (screen) assay and the other
using a phospholipid-rich reagent
(confirm) assay. This ratio is normal-
ized by dividing by the corresponding ratio for PNP performed in the
same assay. The percentage correction based on the study of Tripodi et
al. (3 ) is the difference between the
test plasma screen normalized by
PNP screen and test plasma confirm
also normalized by PNP confirm.
This difference is expressed as a percentage of the normalized screen
clotting time.
Clotting times for the same batch
of PNP provided data for betweenassay imprecision for over 20 assays
(screen, 0.91%; confirm, 0.88%). The
between-assay imprecision over 10
assays for LA-positive patient
plasma was 1.4% for screen and 1.3%
for confirm. The within-assay imprecision for PNP was 0.34% for screen
and 0.27% for confirm (n ⫽ 12), and
for the LA-positive patient plasma
was 0.61% for screen and 0.59% for
confirm.
For the dRVVT-based assays we
used the La Screen and La Confirm
reagents from Life Therapeutics. All
tests were performed on an STA-R
coagulation analyzer from Diagnostica Stago. We used the ␹2 test to
compare the LA positive pick-up rate
in the LR and percentage correction
assays. We constructed 2 ⫻ 2 tables
to compare results derived from the
LR and from the percentage
correction.
Both LR and percentage correction
for 70 individual normal individuals
were normally distributed. The cutoff values (2 SD above the mean) for
70 normal individuals were LR ⫽
1.08 and percentage correction ⫽
13.5%.
For all patients, including the subset on OAC, the overall agreement of
the 2 methods (LR and percentage
correction) was 462/491 ⫽ 94%. The
increased positive test rate for percentage correction (96/491 ⫽ 19.5%)
compared with the LR (73/491 ⫽
14.9%) was statistically significant at
P ⫽ 0.05 (Table 1).
We subdivided the patient population into those on OAC (n ⫽ 85)
and those not on OAC (n ⫽ 406). For
patients on OAC, the concordance
between the 2 methods was 74/85 ⫽
87%. The increased positive test rate
for percentage correction (23/85 ⫽
27%) and LR (14/85 ⫽ 16.9%) did not
differ significantly.
Samples from 26 patients (16
women, ages 19 –76 years, median
age 47 years) were LA positive by
percentage correction and LA negative by LR (Table 1). The incidence of
thrombotic disease (17 venous
thromboembolic and 2 arterial
thromboembolic) was high (73%) in
this group. Six of these patients, all of
whom were diagnosed with venous
thromboembolic
disease,
were
clearly positive with a percentage
correction ⬎3 SD (⬎18.1%) above the
mean.
It has been reported (6, 7 ) that 1:1
mixes of test plasma and PNP may
dilute a weak, yet potentially pathogenic LA, yielding a false-negative
result. The high incidence of thrombotic disease in the 26 patients (9 of
whom were on OAC) poses the question of whether LA activity was diluted by the 1:1 mix in these patients.
The increased sensitivity of the percentage correction to the presence of
a potential LA compared with the LR
may well be of clinical value, warranting further studies to clearly
characterize the presence of an antiphospholipid antibody in these
patients.
Table 1. Results of all 491 patients comparing the 2 methods of data analysis,
dRVVT and percentage correction.
Negative test, ⬍1.07
Positive test, ⬎1.07
Total
Negative Test
(<13.53)
Positive Test
(>13.53)
Total
392
3
395
26
70
96
418
73
491
A direct comparison of the dRVVT LR and percentage correction shows an agreement of 462/491 (94.1%).
Of the discrepant results (n ⫽ 29), 26 patients, 19 with clearly established thrombotic disease, were positive
for LA by the percentage correction method only, and the remaining 3 patients with no evidence of thrombotic
disease were positive by the LR method only. The percentage correction is significantly more sensitive to the
presence of LA in the overall patient population (P ⫽ 0.05).
References
1. Brandt JT, Triplett DA, Alving B, Scharrer I.
Criteria for the diagnosis of lupus anticoagulants: an update. On behalf of the Subcommittee
on Lupus Anticoagulant/Antiphospholipid Antibody of the Scientific and Standardisation Committee of the ISTH. Thromb Haemost 1995;74:
1185–90.
2. Jacobsen EM, Barna-Cler L, Taylor JM, Triplett
DA, Wisløff F. The lupus ratio test. An interlaboratory study on the detection of lupus anticoagulants by an APTT-based, integrated and semi-
Clinical Chemistry 53, No. 1, 2007
3.
4.
5.
6.
7.
quantitative test. Thromb Haemost 2000;83:
704 – 8.
Tripodi A, Chantarangkul V, Clerici M, Mannucci
PM. Laboratory diagnosis of Lupus Anticoagulants for patients on oral anticoagulant treatment. Thromb Haemost 2002;88:583– 6.
Jennings I, Greaves M, Mackie IJ, Kitchen S,
Woods TAL, Preston FE. Lupus anticoagulant
testing: improvements in performance in a UK
NEQAS proficiency testing exercise after dissemination of national guidelines on laboratory
methods. Br J Haematol 2002;119:364 –9.
Aboud MR, Ma DDF. A comparison between two
activated protein C resistance methods as routine diagnostic tests for factor V Leiden mutation. Br J Haematol 1997;97:798 – 801.
Thom J, Ivey L, Eikelboom J. Normal plasma
mixing studies in the laboratory diagnosis of
lupus anticoagulant. J Thromb Haemost 2003;
1:2689 –91.
Jennings I, Woods TAL, Kitchen S, Preston E,
Hughes GRV. Potentially clinically important inaccuracies in testing for the lupus anticoagulant:
an analysis of the results from the surveys of the
UK National External Quality Assurance Scheme
(NEQAS) for blood coagulation. Thromb Haemost
1997;77:934 –7.
Margaret Aboud1,2,3*
Clair Roddie1,2
Christopher Ward1,2,3
Luke Coyle1,2,3
1
2
Department of Haematology
and Transfusion Medicine,
Royal North Shore Hospital
Sydney, 2065, Australi
Northern Blood Research Centre
University of Sydney
RNSH campus
Sydney, Australia
3
Pacific Laboratory
Medicine Services
Royal North Shore Hospital
Sydney, Australia
* Address correspondence to this author at: PaLMS Haematology, Royal
North Shore Hospital, St. Leonards,
NSW, 2065, Australia. Fax 612-9926-6066;
e-mail [email protected].
au.
DOI: 10.1373/clinchem.2006.078683
Upconversion Fluorescence Enables
Homogeneous Immunoassay in
Whole Blood
To the Editor:
Many medical conditions are related to
alterations in protein or hormone con-
145
Fig. 1. Principle (A) and standard curves (B) of the upconversion FRET-based competitive
homogeneous E2 immunoassay.
(A), the anti-E2-Fab-coated FCD-546 –1 UCP (donor) can bind E2 (analyte) and E2-AF680 conjugate (acceptor).
Phosphor emission at 650 – 670 nm excites bound acceptors by FRET under the infrared excitation at 980 nm
and sensitized acceptor emission can be measured at 740 nm. (B), standard curves in buffer (F, wide solid
line), whole blood (f, dashed line) and plasma (Œ, narrow solid line) were obtained after 45 min of incubation
using 15 mg/L UCP donor and 4 nmol/L acceptor conjugate. IC50 values (concentration that inhibited 50% of
the maximum signal) of the assay were 0.72 nmol/L, 1.25 nmol/L, and 1.35 nmol/L, respectively. Standard
curves were fitted to the data based on means of 4 replicates using the program Origin 6.0 (OriginLab
Corporation) and the logistic function y ⫽ (A1 ⫺ A2)/[1 ⫹ (x/x0)p)] ⫹ A2, where A1 and A2 correspond to
maximum and minimum values of the response, respectively, p is the slope and x0 is the IC50 value. The error
bars indicate the SD of the value. cts, counts.
centrations in blood. These changes are
commonly detected with heterogeneous immunoassays that require a
separation step before signal measurement. A simple homogeneous
assay performable directly in whole
blood would be useful in clinical
diagnostics, but the performance of
homogeneous
fluorescence-based
immunoassays has been severely
limited by autofluorescence and
strong absorption of ultraviolet and
visible light in whole blood. These
problems might be avoided by the
use of near-infrared excited upconversion fluorescence. We recently described a novel homogeneous assay
method based on upconversion fluorescence resonance energy transfer
(FRET). In this method an upconverting phosphor (UCP) is used as a
donor and a fluorescent protein or a
small-molecular fluorescent dye is
used as an acceptor (1, 2 ). We report
here the use of a competitive homogeneous immunoassay for 17␤-estradiol (E2) in whole blood as a model
to demonstrate the application of upconversion FRET using a near-infrared fluorescent acceptor dye (Fig. 1A;
and see Fig. 1 in the Data Supplement that accompanies the online
version of this Letter at http://www.
clinchem.org/content/vol53/issue1).
A UCP (Luminophor SPF) with a
structural composition of La2O2S:
Yb3⫹,Er3⫹ was coated as described
earlier (1–3 ) with an E2-specific recombinant antibody Fab fragment
with a known cross-reactivity profile
(4, 5 ). A Fab fragment was used to
provide more favorable donor–acceptor distances for resonance energy transfer (5 ). The average size of
the UCP donor particle was ⬃390
nm, determined with a Coulter N4
Plus submicron particle size analyzer
(Beckman Coulter). A succinimidyl
ester of a small-molecular acceptor
dye, Alexa Fluor 680 (AF680) (Molecular Probes, Invitrogen Corp.), was
coupled with a 2.5-fold molar excess
of the amino-derivative of E2 (6-oxoestradiol 6-[O-(6-aminohexyl)oxime])
as described (2 ). Whole blood was
collected in lithium heparin anticoagulated tubes (Venoject; Terumo
Europe) from male volunteers at the
Department of Biotechnology, University of Turku, and used in the
assay with their informed consent.
The E2 immunoassay was carried out
in buffer, plasma, and blood according to an assay principle previously
described (2 ). Blood and plasma
samples were diluted in buffer to
comprise 20% of the total 45-␮L reaction volume. The reactions were